Re: [cocci] behavior change in semantic patches since c36b51ffc889 ("don't drop down on single statement")

[Jacob Keller <jacob.e.keller@intel.com>]

On 3/6/2025 4:55 PM, Julia Lawall wrote:
> 
> 
> On Thu, 6 Mar 2025, Jacob Keller wrote:
> 
>> Hi Julia,
>>
>> We use semantic patches for code cleanup regularly, and recently noticed
>> a change in behavior some some semantic patches that are run as part of
>> a build process. I've provided the patch and an explanation of my
>> discoveries thus far. Hopefully help track down whats wrong and whether
>> we need to update these patches, there is some other workaround we can
>> do, or there is actually a bug in coccinelle/spatch.
> 
> Thanks for the report, and espcially for the bisect.  I don't think it
> would be good to just revert the commit, since it is solving a real
> problem.  I will see how to get your cases working properly.
> 

Thanks :)


>>> @@
>>> expression hw, ptr;
>>> @@
>>> (
>>> - ice_free(hw, ptr);
>>> + devm_kfree(ice_hw_to_dev(hw),ptr);
>>> )
> 
> This doesn't need the starting and ending ()
> This is just a comemnt, it won't solve the problem.
> 
> julia
> 

Yep. I am honestly not sure where those came from, we've had this one
around for some time.